Digital Media Art Creation and Expression Modelling Based on Virtual Reality Technology

Abstract

Art enthusiasts may now find, learn about, and enjoy more works of many creative genres thanks to the advancement of Internet digital media technology. Different artistic painting styles inspire artists to come up with fresh ideas, which leads to a variety of artistic creation styles combined with various artistic creation approaches. In this manuscript, Dual-path Multi-scale Attention Guided network based on Media Creation using Virtual Reality technology (MCVR- DPMSAGN- EHOA) is proposed. The input data are collected from MNIST dataset. Then the data are pre-processing using Fast Resampled Iterative Filtering (FRIF) to remove noise data. The pre-processed data are given into Quadratic Phase Wavelet Packet Transform (QPWPT) for extracting the features such as Human Gesture and Posture image to enhance the image pixel sorting of the filter window. The features are given into DPMSAGN for classification of textures, lighting, object details, and visual styles. In general, the DPMSAGN does no express adapting optimization strategies to determine optimal parameters to ensure accurate classification. Therefore, the Elk Herd Optimization Algorithm (EHOA) is introduced for optimizing DPMSAGN. The MCVR- DPMSAGN- EHOA proposed is implemented in Python working platform. Performance parameters including precision, accuracy, F-score, computing time, specificity, and sensitivity are measured and evaluated in the proposed technique. The proposed method attains 25% accuracy, 18.77sec for lower Computation time, 21.87% for higher F1 score and 19% for higher specificity than the existing methods like enhanced Media Creation using Virtual Reality through Artificial Neural Network (MCVR-ANN), Media Creation using Virtual Reality through Circle Search Algorithm (MCVR-CSA) and Media Creation using Virtual Reality through Skill Optimization Algorithm (MCVR-SOA) respectively.